Y
HN Search
Hacker News Search
new
|
comments
|
top
|
jobs
henning
searching PlanetScale…
1.
▲
2.
▲
3.
▲
4.
▲
5.
▲
6.
▲
36 ms
·
421.
▲
A Compilers View of OpenMP
(youtube.com)
6 points
by
henning
5y ago
|
0 comments
422.
▲
by
henning
5y ago
You can have a good career and a good life without being a macho asshole bully. Grow up.
423.
▲
Sudo on Linux Was Hacked
(youtube.com)
2 points
by
henning
5y ago
|
1 comments
424.
▲
PCG32: The Perfect PRNG for Roguelikes (2018)
(steveasleep.com)
62 points
by
henning
5y ago
|
38 comments
425.
▲
Shitty First Startups (2015)
(youtube.com)
2 points
by
henning
5y ago
|
0 comments
426.
▲
Tilt and Residual Tilt – How to Frame Frustrating Situations
(youtube.com)
2 points
by
henning
6y ago
|
0 comments
427.
▲
by
henning
6y ago
The desire to go "truly serverless" and pretend the computer does not actually exist is absolutely delusional. The refusal to acknowledge that software will never be anything beyond executable data on some kind of computer, somewh
428.
▲
Reconnecting 170k Phone Customers in NYC After a Major Fire (1975)
(youtube.com)
3 points
by
henning
6y ago
|
0 comments
429.
▲
The Edges of Our Universe
(arxiv.org)
107 points
by
henning
6y ago
|
49 comments
430.
▲
NNCP: Lossless Data Compression with Neural Networks
(bellard.org)
2 points
by
henning
6y ago
|
0 comments
431.
▲
Chasing Top Fund Managers
(youtube.com)
1 points
by
henning
6y ago
|
0 comments
432.
▲
by
henning
6y ago
Are Flask/Django APIs that take floats as parameters vulnerable to this under any circumstances? Any POC code that can be used for testing?
433.
▲
The Factory Must Grow: Automation in Factorio
(arxiv.org)
1 points
by
henning
6y ago
|
0 comments
434.
▲
What Compilers Can and Cannot Do
(youtube.com)
1 points
by
henning
6y ago
|
0 comments
435.
▲
Warp vs. Actix-Web
(youtube.com)
2 points
by
henning
6y ago
|
0 comments
436.
▲
by
henning
6y ago
It matters if you're applying for a job and the people interviewing or the hiring manager think you should. People like Mike Acton and Jonathan Blow give incredibly condescending speeches about "practice" like you should be a
437.
▲
by
henning
6y ago
I'd rather hold banks, regulators, and governments accountable so that I can safely and sensibly transfer funds rather than having to rely on bullshit nerd Ponzi schemes.
438.
▲
by
henning
6y ago
I am continuing to play around with genetic programming in Rust. Specifically, I am working towards building my own implementation of a recent team-based hierarchical algorithm called Tangled Program Graphs. http://stephenkelly.c
439.
▲
The code behind Quake's movement tricks explained
(youtube.com)
3 points
by
henning
6y ago
|
0 comments
440.
▲
by
henning
6y ago
I'd try to build something real on it while doing as much from scratch as possible. AWS has so much crap in it that no one can know everything. I'd focus on the major, most popular services.
441.
▲
Less Is More: Exploiting Standard Compiler Optimizations for Better Performance
(arxiv.org)
2 points
by
henning
6y ago
|
0 comments
442.
▲
by
henning
6y ago
I don't do much C programming, but I set up a benchmark of the code in this post and it does indeed seem to be about 25% faster for the toy instruction set in the post. https://github.com/fearofcode/labeled_goto_te
443.
▲
by
henning
6y ago
Strange, I use arrow functions and const to make variables immutable all the time. Classes are widely used in React although with the arrival of hooks they aren't as necessary. I use async functions and generators extensively with redu
444.
▲
Computed goto for efficient dispatch tables (2012)
(eli.thegreenplace.net)
47 points
by
henning
6y ago
|
26 comments
445.
▲
Text Files Are Underestimated
(borjaportugal.com)
3 points
by
henning
6y ago
|
0 comments
446.
▲
The Future of RTS
(docs.google.com)
189 points
by
henning
6y ago
|
271 comments
447.
▲
Fuzzy logic rice cookers (2016)
(fivethirtyeight.com)
64 points
by
henning
6y ago
|
81 comments
448.
▲
Teaching a computer to strafe jump in Quake with reinforcement learning
(youtube.com)
2 points
by
henning
6y ago
|
0 comments
449.
▲
by
henning
6y ago
As someone who has played with writing trading bots but never traded them with real money, some advice: if your results seem too good to be true, they probably are. Your trading bot may be doing unrealistic things or its results may not be
450.
▲
Better Quake strafe-jumping with genetic algorithms [video]
(youtube.com)
3 points
by
henning
6y ago
|
1 comments
More ›